This document summarizes a study that used label-free LC-MS/MS to simultaneously identify and quantify proteins in E. coli grown in different carbon sources (glucose, lactose, acetate). The methodology involved growing E. coli in different media, extracting and digesting proteins, and performing LC-MS/MS analysis using alternating low and elevated collision energies to obtain precursor and fragment ion data. Over 7,000 accurate mass measurements of precursors and 25,000 of fragments were obtained. Multiple peptides were identified for relative protein quantitation. Results showed differential expression of proteins involved in carbon utilization pathways between the growth conditions. The label-free quantitation was consistent with other omics methods.

1. Asilomar Qualitative and Quantitative Analysis of E. coli by Time-Resolved Mass Spectrometry
2005 Jeffrey C. Silva, Richard Denny, Craig Dorschel, Guo-Zhong Li, Marc V. Gorenstein, Keith Richardson, Daniel B. Wall, Scott J. Geromanos
Waters Corporation, Core Technology/Proteomics, 34 Maple Street, Milford MA 01757
OVERVIEW RESULTS RESULTS RESULTS RESULTS
Purpose:
?? Use a novel, label-free LCMS methodology to monitor protein levels in complex protein mixtures. Alignment of Precursor and Fragments Peptide Identification by LCMSE Multiple Peptides for Relative Quantitation Patterns in Chaos
?? Use E. coli as a model system to study proteins influenced by carbon source utilization. Citrate synthase, GLTA (1937.8519 MH+) Succinyl CoA synthase (beta), SUCC (1556.8199 MH+) ACEA ALDA TUFA GALE ACEA ALDA TUFA GALE
?? Illustrate the ability to simultaneously identify and quantify proteins among complex samples.
A Precursor
A B
INTRODUCTION y13
Profiling of expressed microbial proteins has diverse applications ranging from the response of bacteria to exter-
nal stimuli to the bacterial and fungal response to pharmaceuticals and disinfectants. Separation methods have been inte-
grated with mass spectrometry to reduce the complexity of biological, digested protein mixtures for efficient monitoring y12
of differential protein expression, but these methods can be time consuming, laborious and not amenable to accurate IDH MDH RL1 RS1
IDH MDH RL1 RS1
quantification. A robust protein profiling method should provide accurate, rapid and reproducible detection of changes in
the level of protein expression or presence of new proteins. The intent of this poster will be to demonstrate a novel, la- Protein chain elongation factor, TUFA (2117.1440 MH+) Phosphoglycerate kinase, PGK (2242.1350 MH+)
bel-free methodology (Figure 1) for reproducibly monitoring protein expression levels and illustrate its ability to simul-
y11
taneously identify proteins/peptides in a complex matrix such as E. coli. C D
y9
METHODS
LC Conditions y8
?? Waters CapLC HPLC System at 5.0 microliters/min flow rate. Figure 7. Differential Expression of Proteins. The multiple peptide measurements to each protein provides a means to
?? AtlantisTM Column (350 ? m X 15 cm, 5 ? m particles). y4 E
Figure 4. Sequence Validated Peptides by LCMS . A PLGS database search against and E. coli protein database obtain a 95% confidence interval for the relative expression level for each binary comparison. The pattern obtained from
Figure 6. Differential Expression of Peptides. Clustering and quantitative analysis of the LCMSE data shows
?? Gradient: 6% to 40% Acetonitrile/0.1% Formic Acid over 100 min. identified the following peptides to: A) citrate synthase (GLTA, GVFTFDPGFTSTASCESK), B) succinyl CoA syn- the relative abundance of each protein from each condition provides a mechanism to group related proteins according to
that identified peptides in each condition corresponding to differentially expressed proteins have expression ratios their response to the applied perturbation. Ribosomal proteins RL1 and RS1 show a pattern across the various condi-
thase (SUCC, GLTDAAQQVVAAVEGK), C) protein elongation factor EF-Tu (TUFA, AIDKPFLLPIEDVFSISGR),
within a narrow range. The relative expression of the protein is determined from these multiple peptide ratios, pro- tions, as do a majority of the other identified ribosomal proteins in this study. Other proteins such as ACEA, ALDA and
D) phosphoglycerate kinase (PGK, ADEQILDIGDASAQELAEILK). The mass error was 3.65, 5.49, 9.58 and 3.49
MS Conditions ppm from theoretical, respectively, for each of the corresponding peptides. MDH share a similar pattern which can be explained by their role in carbon utilization. GALE has a unique pattern
which can be attributed to its role in lactose catabolism.
?? Prototype QTof Ultima mass spectrometer in V-mode (? 12K ? WHM)
~ F
?? A Nano-LockSpray ion source (~ 5ppm mass accuracy). Quantitative Comparison of Peptides/Proteins LACZ
PGI
-0.65 +/- 0.13
?? Alternate scanning acquisition, 1.85 seconds for the low (8 eV) and elevated (28-35 eV) collision energy channels. 1,2 ND
B A StDev = 0.22 B Lactose
Lactose
0.12 +/- 0.15
-0.78 +/- 0.10
Replicate injections
Sample Preparation GALM
1.32 +/- 0.21
2.13 +/- 0.21
-0.99 +/- 0.14
Media and Growth Conditions:
? E. coli (ATCC10798, K-12) GALK
Acetate
? M9 minimal media with either 0.5% glucose, lactose or acetate. Lactose
Protein Extract Preparation: -1.91 +/- 0.35 GAPA
-0.58 +/- 0.09
? Cells were suspended in 5mL/1gm biomass in lysis buffer (50 mM Ammonium Bicarbonate, pH 7.5, 1 mM EDTA).
Replicate injections of acetate condition PGK 0.20 +/- 0.08
? Cells were sonicated using a Microson XL Ultrasonic Cell Disrupter. GALE 0.53 +/- 0.11 -0.77 +/- 0.14
? Debris was removed by centrifugation at 15k rpm for 30 min at 4° C. Figure 3. Alignment of Precursor and Associated Fragments. A) An SIC of a doubly-charged precursor peptide of
1.06 +/- 0.29 0.15 +/- 0.09
LACZ 3.00 +/- 0.29 0.38 +/- 0.07
Protein Digest Preparation: StDev = 0.25
? Protein was reduced (10 mM DTT) and alkylated (30 mM iodoacetamide) in presence of 0.05% Rapigest.
945.664 m/z from the low energy channel (Precursor, function 1) having an apex retention time of 57.14 min. Six SICs
from the elevated energy channel (474.354 m/z, 843.637 m/z, 900.662 m/z, 1114.855 m/z, 1185.882 m/z and 1242.928
C GALT
Lactose/Glucose D -1.99 +/- 0.24
? Protein was digested using modified trypsin (Promega) at a concentration of 50:1 (E. coli protein to trypsin). m/z, function 2) share the same apex retention time of 57.17 min which is within one scan of the highlighted precursor,
GALD
GALM
? Tryptic peptide solution was centrifuged at 13,000 rpm for 10 min before LCMS analysis. B) The elevated energy spectrum (time-resolved mass measurements) associated with the precursor ion described DGAL ENO
above. An E. coli databank search by PLGS of this ion (monoisotopic/lock-mass corrected MH+ = 1890.0109) and its -0.42 +/- 0.09
associated fragments identifies this peptide, GYINSLGALTGGQALQQAK, to isocitrate lyase (ACEA). The observed TUFA 0.08 +/- 0.07
Data Processing mass was determined to be 1.23 ppm less than the theoretical mass. The six SICs illustrated above correspond to the six
corresponding fragment ions to the identified peptide: y4, y8, y9, y11, y12 and y13.
-0.54 +/- 0.05
The Waters Protein Expression Informatics software is used to: Lactose versus glucose
PTA ACKA
? Process raw data: Generate an inventory of peptides from each LCMSE analysis. Time-Resolved Mass Measurements ACEE 0.47 +/- 0.10 0.63 +/- 0.22 A
Protein levels as determined by label-free quantitation from this study.2
? Align peptides: Cluster detected peptides from each sample into a single matrix by accurate mass and retention time. 0.23 +/- 0.08 0.04 +/- 0.10 0.05 +/- 0.09 B
Transcript levels as determined by M. K. Oh etal.3
? Perform label-free quantitation: Normalize all data to peptides of TUFA and determine relative ratios of identical of Precursors and Fragments 0.08 +/- 0.06 0.44 +/- 0.13 0.58 +/- 0.23 C
Protein levels as determined by 2-D gel electrophoresis by L. Peng etal.4
peptides from each condition. E ACEA
Acetate/Glucose StDev = 0.98 F 0.15 +/- 0.07 ND = not determined, Acetate = Unique to acetate condition
? Identify peptides/proteins: Accurate mass measurement of precursor and associated fragments and/or PMF-search ACS
using clustered peptides from the quantitative results. SUCC ACS Table 1. Comparison of Label-Free Relative Quantitation to 2DGE and Transcriptional Profiling.
MDH 2.81 +/- 0.25
SUCD 1.98 +/- 0.11
ND
-0.17 +/- 0.04
ND
Alternate Scanning LCMS: LCMSE
THRC 2.15 +/- 0.11
ACEB
GLTA
2.32 +/- 0.26
-0.09 +/- 0.10
CONCLUSIONS
3.35 +/- 0.41
? A parallel LCMS method (LCMSE) was used to simultaneously quan-
ND 2.43 +/- 0.25
Acetate versus glucose ND
ACEA
4.48 +/- 0.27
ACNA
0.27
tify and identify proteins involved in carbon source utilization.
? The parallel LCMS method provides a means to obtain time-resolved
0.16 +/- 0.15
+/-
4.29 +/- 0.21
Precursors G Acetate/Lactose
ACEA
StDev = 1.01 H FUMC
-0.52 +/- 0.14
0.09
0.05 +/- 0.09 accurate mass measurements for precursors and fragments.
? Protein identifications from several metabolic pathways of E. coli
MDH -0.60 +/- 0.17
GALM were made using the accurate mass measurements of the detected pre-
GALE
cursors and their associated fragments.
Acetate versus lactose ? A label-free quantitation method was used to determine the relative
SUCC
1.88 +/- 0.14
levels of these proteins among the different growth conditions.
Figure 5. Differential Protein Expression. Scatter plots (A, C, E and G) and corresponding histograms (B, D, F and
-0.16 +/- 0.07
2.03 +/- 0.13 SUCD SUCA ? Results are consistent with those obtained by other methods.3,4
H) of the natural logarithm of the relative intensity ratios of matched peptides. Multiple peptides corresponding to a 1.82 +/- 0.14 1.99 +/- 0.19
-0.14 +/- 0.06 -0.17 +/- 0.09
Fragments subset of identified proteins are highlighted in the scatter plot for each binary comparison. The standard deviation for 1.96 +/- 0.14 2.16 +/- 0.27
the ln(intensity ratio) of the matched peptides are indicated in the histogram plot for each binary comparison. A) The ln
(intensity ratio) of matched peptides between two replicate injections of detected peptides from E. coli grown on ace- Figure 8. Glycolysis, Citric Acid Cycle and Glyoxylate Shunt. The relative quantitation of each protein is color-
REFERENCES
tate, C) Lactose versus glucose: red-LACZ, blue-GALT, yellow-GALD, black-GALM, green-DGAL and pink-TUFA), coded for each binary comparison. The following color-coding scheme is used: Acetate versus Glucose (red), Lactose 1
R. H. Bateman etal. (2002) U. K. Patent 2,385,918A.
Figure 2. Ion Detections from an LCMSE Analysis of E. coli. A total of 7147 and 25273 accurate mass measure- E) Acetate versus glucose: green-ACEA, red-SUCC, blue-SUCD, yellow-ACS, black-THRC, G) Acetate versus lac- versus Glucose (blue) and Acetate versus Lactose (black). The relative quantitation is reported as: ln(ratio) +/- 95% 2
J. C. Silva etal. Anal. Chem. (2004) 77, 2187-2220.
ments from the low and elevated energy channels, respectively. The six highlighted precursors (blue) are identified to tose: red-GALE, blue-GALM, yellow-ACEA, black-MDH. Unique peptides to each condition is indicated at the ex- Confidence Interval. ND = not determined, Lactose = unique to lactose, Acetate = unique to acetate, Glucose = unique 3
Figure 1. LCMSE Acquisition. Simultaneous acquisition of precursor and fragment ions for label-free quantitation and ribosomal protein RL9 and constitute approximately 45% protein sequence coverage. The fragmentation data obtained treme values in both the scatter plots and histogram plots (top or left = numerator condition, bottom or right = denomi-
M. K. Oh etal. J. Biol. Chem. (2002) 277, 13175-13183.
to glucose 4
L. Peng etal. Appl. Microbiol. Biotechnol. (2003) 61, 163-178.
identification of peptides and proteins. from each precursor is highlighted in red. nator condition.).
720001406EN